DESCRIPTION: (Applicant's Abstract) The applicant notes that a vast majority of the established human brain tumor cell lines, human brain tumor explant cells, and more recently, brain tumor tissues examined in situ, over-express receptor for a hemopoietic growth factor human interleukin 13 (hIL-l3). hIL-l3 is homologous to hIL-4. The number of IL-13 binding sites in a primary culture of human glioblastoma multiforme (GBM) cells can be as high as 500,000 per cell. Chimeric cytotoxic proteins composed of hIL-l3 and derivatives of Pseudomonas exotoxin A or Diphtheria toxin were made. The chimeric cytotoxins kill brain tumor cells very potently and their prominent antitumor activities were demonstrated in vivo. These results are of particular interest, since the prognosis of brain cancers remains poor and specific targets for their treatment and new forms of treatment are needed. The receptor for IL-13 that is over-expressed on human brain tumors is different from the one present on normal cells, because it is not shared with IL-4. The applicant proposes to (i) clone the IL-13 binding protein that has been isolated from human malignant gliomas, (ii) document the biodistribution of hIL-l3 binding sites using autoradiography, immunohistochemistry, and quantitative RT-PCR, and (iii) explore a unique possibility to create pharmacologically a tumor-specific hIL-l3 receptor. An approach has been designed to minimize/eliminate normal tissue reactivity of the targeted therapeutics even though the first generation of therapeutics is already very selective. Preclinical evaluation of the cytotoxins will be conducted in rat models of GBM. It is expected that the proposed experiments will lead to further preclinical and clinical trial(s).